Lumbar Neuraxial Ultrasound for Spinal and Epidural Anesthesia a Systematic Review and Meta-Analysis

SPECIAL ARTICLE

Lumbar Neuraxial Ultrasound for Spinal and Epidural Anesthesia A Systematic Review and Meta-Analysis

Anahi Perlas, MD, FRCPC,*† Luis E. Chaparro, MD,‡ and Ki Jinn Chin, MD, FRCPC*†

pinal anesthesia and lumbar epidural anesthesia are commonly Background: This systematic review examines the evidence for prepro- Sperformed anesthetic and/or analgesic techniques with a long cedural neuraxial ultrasound as an adjunct to lumbar spinal and epidural track record of efficacy and safety. However, neuraxial blocks anesthesia in adults. can occasionally be challenging to perform, particularly if the Methods: We searched MEDLINE, EMBASE, and Cochrane Central spinal anatomy is altered or obscured by factors such as obesity, Register of Controlled Trials databases from inception to June 30, 2014, spinal deformities, or previous spine surgery.1 Technical difficulty for randomized controlled trials (RCTs) and cohort studies that reported can result in procedural failure, suboptimal epidural analgesia, data answering one or more of the following 3 questions: (1) Does ultra- and increased needle trauma. It may also increase the risk of both ound accurately identify a given lumbar intervertebral space? (2) Does minor complications such as postdural puncture headache and ultrasound accurately predict the needle insertion depth required to reach backache and major complications including epidural hematoma the epidural or intrathecal space? (3) Does ultrasound improve the effi- and spinal cord injury.2–4 cacy and safety of spinal or lumbar epidural anesthesia? Neuraxial ultrasound is a recent development in the field of Results: Thirty-one clinical trials and 1 meta-analysis were included in regional anesthesia. A “pre-procedural” ultrasound examination this review. Data from 8 studies indicate that neuraxial ultrasound can iden- of the spine accurately delineates the underlying relevant anatomy, tify a given lumbar intervertebral space more accurately than by landmark thus aiding in successful insertion of a spinal or epidural needle; palpation alone. Thirteen studies reported an excellent correlation between this has also been termed “ultrasound-assisted ” neuraxial blockade. ultrasound-measured depth and needle insertion depth to the epidural or Although real-time ultrasound-guided spinal and epidural intrathecal space. The mean difference between the 2 measurements was techniques have been described, they are distinctly different from within 3 mm in most studies. Thirteen RCTs, 5 cohort studies, and 1 meta- the ultrasound-assisted approach. They remain experimental at analysis reported data on efficacy and safety outcomes. Results consistently this time and will not be discussed in this review. showed that ultrasound resulted in increased success and ease of per- The objective of this review was to examine the evidence formance. Ultrasound seemed to reduce the risk of traumatic procedures supporting the use of preprocedural neuraxial ultrasound to but there was otherwise insufficient evidence to conclude if it significantly facilitate spinal or lumbar epidural anesthesia and, based on improves safety. this, to set forth recommendations for practice. We addressed Conclusions: There is significant evidence supporting the role of neur- 3 distinct clinical questions: axial ultrasound in improving the precision and efficacy of neuraxial anesthetic techniques. 1) Does neuraxial ultrasound accurately identify a given lumbar What's New: We know that neuraxial ultrasound is a useful complement to clinical examination when performing lumbar central intervertebral space? 2) Does neuraxial ultrasound accurately predict the needle neuraxial blocks. It provides anatomical information including the depth of the epidural space, the identity of a given intervertebral level, and the insertion depth required to reach the epidural or intrathecal location of the midline and interspinous/interlaminar spaces. This infor- space? mation can be used to successfully guide subsequent needle insertion. 3) Does neuraxial ultrasound improve the efficacy and safety of spinal or epidural anesthesia? Since 2010, new data from RCTs and 1 meta-analysis suggest that neuraxial ultrasound increases the success and reduces the technical difficulty of lumbar central neuraxial blocks. Findings from the meta- analysis suggest that neuraxial ultrasound reduces the risk of traumatic METHODS procedures, and thus may possibly contribute to the safety of lumbar For this review, we included all randomized controlled trials central neuraxial blocks. (RCTs) and cohort studies involving neuraxial ultrasound and (Reg Anesth Pain Med 2016;41: 251–260) spinal or lumbar epidural anesthesia/analgesia in adult patients. We also included studies involving diagnostic lumbar puncture, given that the needle insertion technique is identical to that of From the *Department of Anesthesia and Pain Management, Toronto Western spinal anesthesia. We excluded studies of real-time ultrasound- Hospital, University Health Network; †Department of Anesthesia, University guided neuraxial blocks as well as those related to interventional of Toronto, Toronto, Ontario, Canada; and ‡Department of Anesthesiology, Hospital Pablo Tobón Uribe, Medellín, Colombia. pain procedures on the spine. Studies that did not report outcomes Accepted for publication October 10, 2014. related to the 3 primary questions were excluded. We performed Address correspondence to: Anahi Perlas, MD, FRCPC, University of Toronto, a literature search of the MEDLINE, EMBASE, and Cochrane Mc Laughlin Pavilion 2-405, 399 Bathurst St, Toronto, Ontario, Canada Central Register of Controlled Trials databases from the time of M5T 2S8 (e‐mail: [email protected]). The authors declare no conflict of interest. inception until June 30, 2014. The following search terms were Drs Anahi Perlas and Ki Jinn Chin received academic time support through a used: ultrasound, ultrasonography, epidural, peridural, subarachnoid Research Merit Award 2011–2013 and 2013–2015, respectively, from the space, epidural analgesia, epidural anesthesia, spinal anesthesia, and Department of Anesthesia, University of Toronto. conduction anesthesia. No language restrictions were applied. The Copyright © 2014 by American Society of Regional Anesthesia and Pain Medicine abstracts of all references identified by the search were independently ISSN: 1098-7339 reviewed by 2 authors. Full text copies of potentially relevant DOI: 10.1097/AAP.0000000000000184 studies were obtained and again underwent independent review

Regional Anesthesia and Pain Medicine • Volume 41, Number 2, March-April 2016 251

Copyright © 2016 American Society of Regional Anesthesia and Pain Medicine. Unauthorized reproduction of this article is prohibited. Perlas et al Regional Anesthesia and Pain Medicine • Volume 41, Number 2, March-April 2016

by 2 authors. Data from studies that met the inclusion criteria RESULTS were entered into a standardized data extraction form. All dis- Seven hundred six citations were identified in the initial agreements were resolved by discussion and mutual consensus search of which 57 were selected as potentially relevant and among the 3 authors of this review. We performed a risk of bias 5 underwent full-text review (Fig. 1). Of these, we excluded 25 studies assessment for each study. The QUADAS-2 tool was used for for the following reasons: 9 were narrative reviews or descrip- studies of diagnostic accuracy of neuraxial ultrasound in tive articles, 6 were in the pediatric population, 6 involved in- identifying lumbar intervertebral spaces. We used the Jadad terventions on the thoracic or cervical spine, and 4 were case score6 and the Cochrane Collaboration's risk of bias assessment 7 reports. One meta-analysis was identified and included in this tool for RCTs looking at the effect of neuraxial ultrasound on review.10 clinical outcomes.7 We performed meta-analysis using RevMan 5.3 (The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, 2014). The treatment effect was expressed as risk Does Neuraxial Ultrasound Accurately Identify a ratio for dichotomous outcomes and as mean difference (MD) Given Lumbar Interspace? for continuous data outcomes, respectively, along with 95% Eight studies11–18 involving a total of 624 patients addressed confidence intervals (CIs). Statistical heterogeneity was assessed this question (Table 1). All 8 studies used a “counting-up” approach using both the χ2 test and I2. The studies that reported on the in which the ultrasound probe was placed in a longitudinal orien- correlation of the ultrasound-determined depth of the epidural tation over the sacrum (identified as a continuous hyperechoic line) or intrathecal space versus needle insertion depth under- and then moved cephalad to identify successive spinous pro- went meta-analysis using R software (version 2.15.3). Fisher cesses or laminae and the corresponding interspinous or inter- z-transformation was applied to the Pearson product moment laminar spaces.19 A low-frequency curved-array probe was used correlation coefficients before meta-analysis. The final results in all studies except one.16 These 7 studies were generally of good of pooled z scores were back-transformed to the pooled correlation quality according to the QUADAS-risk of bias assessment tool coefficients. for diagnostic studies, with only 2 studies receiving a “high” Summary recommendations follow the format suggested rating in one domain each. Five studies examined the agree- by the US Department of Health and Human Services Agency ment between ultrasound and palpation of surface landmarks for Health Care Policy and Research.8 We followed the reporting in identifying a given intervertebral space.13–15,17,18 None of recommendations of the PRISMA statement.9 these studies, however, verified accuracy against a more

FIGURE 1. Flow chart of database search and study selection.

Copyright © 2016 American Society of Regional Anesthesia and Pain Medicine. Unauthorized reproduction of this article is prohibited. Regional Anesthesia and Pain Medicine • Volume 41, Number 2, March-April 2016 Lumbar Neuraxial Ultrasound

TABLE 1. Studies Reporting Accuracy of Intervertebral Space Identification

Country Separate Reference Primary Main Author Year of Origin n Standard Outcome Findings Furness 2002 United Kingdom 50 X-ray Accuracy L3-4 Ultrasound 71% accurate Palpation 30% accurate Duniec 2013 Poland 122 Palpation Agreement 64% agreement 18% higher by palpation by1level 0.8% higher by palpation by2levels 16.4% lower by palpation by1level 0.8% lower by palpation by2levels Halpern 2010 Canada 74 CT scan CUSUM analysis for 90% accuracy 90% accuracy after 22 and 36 scans Lee 2011 United States 51 Palpation Agreement 14% agreement 23% higher by palpation by1level 25% higher by palpation by >1 level Locks 2010 Brazil 90 Palpation Agreement at L3-4 50% agreement Schlotterbeck 2008 United Kingdom 99 Palpation Agreement 36% agreement 50% higher by palpation 14% lower by palpation Watson 2001 United Kingdom 17 MRI US accuracy at L3-4 76.5% accuracy (13/17) 23.5% off by 1 level Whitty 2008 Canada 121 Palpation Agreement 55% agreement 32% higher by palpation established gold-standard imaging modality (the reference Finally, in a learning curve study that used CT as a reference standard). Schlotterbeck et al15 and Whitty et al17 both studied a standard, Halpern et al12 reported an overall identification retrospective cohort of patients who had received labor epidural accuracy rate for ultrasound of 68%. However, analysis of the analgesia and in whom the documented level of epidural learning curve showed that the 2 anesthesiologists in the insertion was correlated with ultrasonographic identification of study with no previous experience with neuraxial ultrasound the intervertebral level corresponding to the visible skin puncture achieved accuracy rates of 90% or greater after 22 and 36 site. They excluded patients with multiple puncture marks or who procedures, respectively. had inadequate documentation of epidural insertion site. For these reasons, they received a “high” risk of bias rating in the “flow and Recommendation ” timing domain of the QUADAS-2 tool. In all 5 studies, the There are consistent data (evidence level IIa) to suggest that agreement between ultrasound and palpation-determined neuraxial ultrasound identifies lumbar intervertebral levels, with interspaces was generally poor with rates ranging from 14% greater accuracy than palpation of surface anatomical landmarks to 64%. In cases of disagreement, palpation-determined (grade B recommendation). landmarks were usually higher than ultrasound-determined – landmarks (52% 78% of cases) and often erred by more than Does Neuraxial Ultrasound Accurately Predict the one interspace. Needle Insertion Depth Required to Reach the The remaining 3 studies used x-ray, magnetic resonance imaging (MRI), or computed tomographic (CT) scan as a separate Epidural or Intrathecal Space? reference standard to verify intervertebral level.11,12,16 Using plain Thirteen studies involving a total of 875 patients examined x-ray of the lumbar spine as a reference standard, Furness et al11 the correlation between ultrasound-measured depth and actual demonstrated that ultrasound correctly identified individual inter- needle insertion depth required to reach the epidural or intrathecal spaces (from L2-3 to L4-5) 71% of the time, whereas palpation space (Table 2).20–32 Nine studies were performed in obstetric was only correct 29% of the time. Furthermore, the margin of patients,20,21,23,24,27–29,31,32 3innon–obstetric surgical patients error never exceeded one level with ultrasound, but was up to 2 (urology, vascular, and orthopedics),22,26,30 and 1 in patients spaces higher or lower in 27% of palpation assessments. These requiring a diagnostic lumbar puncture in the emergency findings are consistent with those reported by Watson et al16 department.25 The quality of the studies was generally good, who, using MRI as their reference standard, found that with the most common deficiency being unclear patient ultrasound accurately identified the L3-4 interspace in 76% selection criteria. Two studies30,32 received a “high” risk of of cases with a margin of error that did not exceed one level. bias rating in the reference standard domain because of lack

TABLE 2. Studies Reporting Accuracy of Ultrasound Measurement of Epidural Space Depth

Bland-Altman Analysis LOA Country Sample Pearson Mean, SD, (95% CI), Author Year of Origin Size (n) CC (r) mm mm mm Patients Structure Evaluated Arzola 2007 Canada 61 0.88 0.1 3.5 −6.6 to 6.9 Obstetrics LF-D complex Balki 2009 Canada 48 0.84 3.0 −7.0 to 13.0 Obstetrics LF-D complex Chin 2009 Canada 50 0.82 2.1 −8.5 to 12.7 Orthopedics LF-D complex and PVB Cork 1980 United States 36 0.98 Obstetrics Lamina Currie 1984 United Kingdom 75 0.96 Obstetrics Lamina Ferre 2009 United States 39 0.80 ER patients for LPs LF-D complex Gnaho 2012 France 31 0.98 2.2 1.8 −1.4 to 5.8 Orthopedics LF-D complex Grau AAS 2001 Germany 36 0.93 7.9 Obstetrics LF-D complex Grau RAPM 2001 Germany 80 0.96 2.0 2.3 −3 to 7 Obstetrics LF-D complex Grau 2002 Germany 150 0.91 1.7 −6.0 to 8.0 Obstetrics LF-D complex Helayel 2010 Brazil 60 0.66 0.04 0.1 −2.3 to 2.2 Orthopedics, urology, LF-D complex vascular Tran 2009 Canada 20 0.89 −4.8 −14.8 to 5.2 Obstetrics LF-D complex Vallejo 2010 United States 189 0.91 Obstetrics LF-D complex

of blinding (ie, the anesthesiologist performing the epidu- approach was used for needle insertion in all studies. The ral procedure was aware of the ultrasound-measured depth to ultrasound landmark used for measuring depth to the epidural the epidural space) (Table 3). The Pearson correlation coeffi- space in most studies was the ventral aspect of the hyperechoic cient reported by the individual studies ranged from 0.66 to ligamentum flavum-dura mater complex (Table 2). The older 0.98.23,26,30 The pooled Pearson product moment correlation studies by Cork et al and Currie measured depth to the ventral coefficient was 0.91 (95% CI, 0.87–0.94), using a random- surface of the laminae; this choice, however, reflects the effects model to account for heterogeneity, suggesting the technological limitations of ultrasound visualization at the time. ultrasound-measured depth of the epidural space was highly Despite these minor variations in method, there was excellent correlated with the depth of the epidural space measured during correlation between ultrasound-measured depth and actual needle the epidural needle insertion (Fig. 2). Four studies measured insertion depth in all studies. It should be noted, however, that a depth to the epidural space using a longitudinal parasagittal strong linear correlation does not necessarily imply accuracy. oblique ultrasound view23–25,31; 3 of these studies also used a To evaluate the accuracy of the ultrasound measurement, 8 of linear-array probe.23–25 All other studies used a low-frequency the more recent studies also performed a Bland-Altman curved-array probe and measured depth to the epidural/intrathecal analysis to study the extent to which the 2 depth measurements space in the transverse midline ultrasound view. A midline differed.20–23,28–31 The ultrasound-determined depth of the

TABLE 3. Risk of Bias Assessment of Studies Reporting the Accuracy of Epidural Space Depth

Risk of Bias Applicability Concerns Country Patient Index Reference Flow and Patient Index Reference Author Year of Origin Selection Test Standard Timing Selection Test Standard Arzola 2007 Canada Low Low Low Low Low Low Low Balki 2009 Canada Low Low Low Low Low Low Low Chin 2009 Canada Low Low High Low Low Low Low Cork 1980 United States Unclear Low Low Low Low Low Low Currie 1984 United Kingdom Unclear Low Low Low Low Unclear Low Ferre 2009 United States Low Low Low Low Low Low Low Gnaho 2012 France Unclear Low Low Low Low Low Low Grau AAS 2001 Germany Low Low Unclear Low Low Low Low Grau RAPM 2001 Germany Unclear Low Low Low Low Low Low Grau 2002 Germany Unclear Low Low Low Low Low Low Helayel 2010 Brazil Unclear Low High Low Low Low Low Tran 2009 Canada Unclear Low Low Low Low Low Low Vallejo 2010 United States Low Low High Low Low Low Low Risk of bias assessment as per the Cochrane Collaboration's tool.7

FIGURE 2. Meta-analysis of studies reporting the correlation of ultrasound-measured versus needle depth of the epidural or intrathecal space. epidural or intrathecal space was found to be accurate within Safety outcomes were consistently reported as secondary 1 to 13 mm of actual needle insertion depth, with 7 of the 8 outcome measures; thus, none of the individual studies were studies reporting an MD of less than or equal to 3 mm. The designed or sufficiently powered to study these outcomes. Four tendency in most studies was for ultrasound to underestimate studies reported a nonsignificant trend toward a lower incidence needle insertion depth; this has been attributed to tissue of headache and backache favoring ultrasound.27,29,34,42 There compression by the probe during the scan. was no difference in the reported rate of unintended dural punctures, which was universally low (<1%).29,32 Only 1 study Recommendation reported a lower incidence of “puncture site hemorrhage” of 36 Data from 13 prospective comparative studies (evidence 7% with ultrasound versus 20% in the control group. No level Ia) consistently show that preprocedure neuraxial ultra- major complications such as epidural hematoma, epidural abscess, sound can be used to accurately predict the needle insertion or intracord injections were reported in any of the RCTs. depth required to reach the epidural or intrathecal space (grade A recent meta-analysis also addressed the question of A recommendation). whether neuraxial ultrasound can reduce the technical failure of lumbar puncture or epidural catheterization.10 The studies Does Neuraxial Ultrasound Improve the Efficacy included were heterogeneous (both preprocedure as well as or Safety of Neuraxial Techniques? real-time ultrasound guidance, and both adult and pediatric 27–29,32–42 patients were included). Nevertheless, the findings were consistent Fourteen RCTs involving 1768 patients (Table 4) with those of our meta-analysis reported in the present review. and 5 prospective cohort studies involving 227 patients20–22,25,26 27–29,32,34,36,38,42 Pooled data from 12 RCTs showed a 79% reduction in the risk of examined efficacy and safety outcomes. Eight RCTs failed lumbar puncture or epidural catheterization (relative risk, and 2 cohort studies20,21 evaluated epidural analgesia in obstetric – 33,37,39 22,26 0.21; 95% CI, 0.1 0.43, P < 0.001) with neuraxial ultrasound. They patients, whereas 3 RCTs and 2 cohort studies evaluated also found a significant reduction in the number of needle re- spinal anesthesia in orthopedic procedures. The remaining 3 − − 35,40,41 25 directions required for success (MD, 1.00; 95% CI, 1.24 to RCTs and 1 cohort study each evaluated diagnostic lumbar −0.75, P < 001). Pooled data from 5 RCTs showed a 73% reduction punctures by emergency physicians. Three RCTs27,33,36 and 1 21 in the risk of a traumatic procedure (relative risk, 0.27; 95% CI, cohort study enrolled only patients in whom technical 0.11–0.67, P = 0.05), which was defined as visible blood on difficulty was expected due to obesity,21,27,33,36 documented 33 33 aspiration or a fluid red blood cell count above a predetermined lumbar scoliosis, or previous lumbar spine surgery. The threshold. The authors further calculated the number needed risk of bias assessment showed the RCTs to be of reasonable to treat to prevent one procedural failure and one traumatic quality, with the commonest deficiency being lack of blinding procedure as 16 and 17, respectively. of the patient and study personnel (Figs. 3 and 4), a limitation that is often difficult to overcome in procedural studies of this nature. Recommendation Thirteen RCTS that reported the risk of technical failure were Data from 14 RCTs and 2 meta-analysis (this article and meta-analyzed (Fig. 5). The combined risk ratio of technical 1 previously published) (level of evidence Ia) support the failure was 0.51 (95% CI, 0.32–0.80) when ultrasound guidance conclusion that neuraxial ultrasound increases the efficacy of was used compared to palpation. In addition, meta-analysis from lumbar epidural or spinal anesthesia by decreasing the risk of 8 RCTs suggests that ultrasound guidance results in a lower technical failure and the number of needle punctures required, number of needle passes required for success (Fig. 6). both in patients with normal surface landmarks and those at

TABLE 4. RCTs Reporting Efficacy and/or Safety Outcomes of Ultrasound-Assisted Neuraxial Techniques

Study Sample Patient Primary Jadad Author Year Country Technique Design Size Population Outcome Secondary Outcomes Score Abdelhamid 2013 Egypt Spinal RCT 90 Adult unspecified First attempt Procedure time, patient 2 success satisfaction Ansari 2014 UAE Spinal RCT 150 OB Procedure time No. needle insertions/passes, 3 headache, backache, patient satisfaction Chin 2011 Canada Spinal RCT 120 Orthopedic First attempt No. needle insertions/passes, 5 difficult spine success failure rate, procedure time Grau AAS 2001 Germany Epidural RCT 72 OB No. punctures, Failure rate, headache, 2 no. levels backache Grau RAPM 2001 Germany CSE RCT 80 OB difficult spine No. punctures, 1 procedure time Grau 2002 Germany Epidural RCT 300 OB Agreement Unintended dural punctures, 2 US-CP complete analgesia Grau 2004 Germany CSE RCT 30 OB No. punctures Procedure time, duration of 2 blockade Lim 2014 Singapore Spinal RCT 170 Non-OB First attempt No. needle redirections, 3 success procedure time, paresthesia, traumatic taps, patient satisfaction Mofidi 2013 Iran LP RCT 80 ER Procedure time No. needle insertions, 2 traumatic taps, pain score Nomura 2007 United States LP RCT 46 ER Success of LP No. attempts 4 Peterson 2014 United States LP RCT 100 ER Success of LP No. needle insertions, traumatic 2 taps, procedure time, pain score, patient satisfaction Sahin 2014 Turkey Spinal RCT 100 OB First attempt No. needle insertions/passes/ 4 success levels attempted, failure rate, procedure time, paresthesia, headache, backache Vallejo 2010 United States Epidural RCT 370 OB Incidence of No. attempts, unintended 3 epidural catheter dural puncture replacement Wang 2012 China CSE RCT 60 OB obese patients First attempt Procedure time, complications, 2 success puncture site hemorrhage

risk of difficult insertion due to obesity, scoliosis, or previous information provided by neuraxial ultrasound results in increased spine surgery (grade A recommendation). efficacy as evidenced by a reduction in the risk of failure and a lower number of needle passes required for success. Epidural hematoma and spinal cord injury due to unintended DISCUSSION intracord injection are rare but serious complications of neuraxial Although the feasibility of neuraxial ultrasound imaging anesthesia.4 Multiple insertion attempts and “traumatic insertion” was first reported several decades ago,23,24 it was not until increase the risk of epidural hematoma4,44 and an inaccurate the early 2000s that the role of neuraxial ultrasound as we assessment of the location of intervertebral spaces can lead to understand it today became established following pioneering unintended intracord injection resulting in spinal cord injury work by Grau et al and significant advances in ultrasound and permanent neurologic sequelae.45,46 Given the very low technology resulting in greater resolution.27–29,34 Since a previous baseline incidence of these catastrophic complications (usually review,43 more data have become available for non–obstetric less than 1 in 100,000 cases), it is not feasible to design prospective patients22,25,26,33,37,39–41 and for patients presumed at risk for studies to conclusively prove that image guidance improves difficult insertion due to obesity, scoliosis, or previous surgery.33 safety. However, the evidence strongly suggests that preprocedure These special patient populations are clinically important neuraxial ultrasound prevents the occurrence of several well- because they are at increased risk for technical difficulty. The recognized mechanisms of injury. present review identified 31 studies that addressed at least 1 By increasing the accuracy of needle placement and de- of 3 driving clinically relevant questions. creasing the number of needle passes, ultrasound may result in Studies evaluating the “diagnostic” performance of ultrasound less traumatic procedures, likely contributing to the prevention as an extension of the physical examination consistently show of epidural hematoma.23–25,29,32 Similarly, by improving the that it enhances the accuracy of landmark identification com- accuracy of intervertebral space identification, a preprocedure pared with palpation of surface landmarks alone, and that it spinal ultrasound could help prevent injuries to the conus accurately measures the depth of the epidural space. A growing medullaris that are consistently associated with a higher-than- body of evidence suggests that the additional anatomical intended needle insertion point resulting from imprecise surface

FIGURE 3. Risk of bias of individual RCTs reporting efficacy and safety outcomes following the Cochrane Risk of Bias assessment tool for RCTs.

FIGURE 4. Summative risk of bias of RCTs reporting efficacy and safety outcomes following the Cochrane Risk of Bias assessment tool for RCTs.

FIGURE 5. Meta-analysis of RCTs (using RevMan 5.3, the Cochrane Collaboration) reporting the risk of technical failure of neuraxial procedures with and without ultrasound imaging.

FIGURE 6. Meta-analysis of RCTs (using RevMan 5.3, the Cochrane Collaboration) reporting the number of needle passes required for neuraxial procedure success with and without ultrasound imaging.

12. Halpern SH, Banerjee A, Stocche R, Glanc P. The use of ultrasound for TABLE 5. Summary Statements, Grades of Recommendations, lumbar spinous process identification: a pilot study. Can J Anaesth.2010; and Supporting Level of Evidence 57:817–822. Grade of Level of 13. Lee AJ, Ranasinghe JS, Chehade JM, et al. Ultrasound assessment of Outcome Recommendation Evidence the vertebral level of the intercristal line in pregnancy. Anesth Analg.2011; 113:559–564. Increased accuracy of BIIa14. Locks Gde F,Almeida MC, Pereira AA. Use of the ultrasound to determine identification of lumbar the level of lumbar puncture in pregnant women. Rev Bras Anestesiol. interspaces 2010;60:13–19. Accurate measurement of the AIa depth of the epidural and 15. Schlotterbeck H, Schaeffer R, Dow WA, Touret Y, Bailey S, Diemunsch P. intrathecal space Ultrasonographic control of the puncture level for lumbar neuraxial block – Improved efficacy of neuraxial AIain obstetric anaesthesia. Br J Anaesth. 2008;100:230 234. anesthesia 16. Watson MJ, Evans S, Thorp JM. Could ultrasonography be used by an Improved safety of neuraxial BIIIanaesthetist to identify a specified lumbar interspace before spinal anesthesia anaesthesia? Br J Anaesth.2003;90:509–511. Following the format suggested by the US Department of Health and 17. Whitty R, Moore M, Macarthur A. Identification of the lumbar interspinous – Human Services Agency for Health Care Policy and Research.8 spaces: palpation versus ultrasound. Anesth Analg.2008;106:538 540, table of contents. 18. Duniec L, Nowakowski P, Kosson D, Lazowski T. Anatomical landmarks 11,12,16,17 landmarks. Therefore, level III evidence supports a based assessment of intravertebral space level for lumbar puncture is grade B recommendation that neuraxial ultrasound may help misleading in more than 30%. Anaesthesiol Intensive Ther. 2013;45:1–6. improve the safety of neuraxial anesthesia (Table 5). 19. Chin KJ, Karmakar MK, Peng P.Ultrasonography of the adult thoracic and lumbar spine for central neuraxial blockade. Anesthesiology. 2011;114: ACKNOWLEDGMENT 1459–1485. The authors thank Ying Qi from Mississauga, Ontario, for 20. Arzola C, Davies S, Rofaeel A, Carvalho JC. Ultrasound using the her support with statistical analysis and meta-analysis of transverse approach to the lumbar spine provides reliable landmarks for correlation studies. labor epidurals. Anesth Analg. 2007;104:1188–1192. REFERENCES 21. Balki M, Lee Y, Halpern S, Carvalho JC. Ultrasound imaging of the lumbar spine in the transverse plane: the correlation between estimated and 1. De Filho GR, Gomes HP, de Fonseca MH, Hoffman JC, Pederneiras SG, actual depth to the epidural space in obese parturients. Anesth Analg.2009; Garcia JH. Predictors of successful neuraxial block: a prospective study. 108:1876–1881. Eur J Anaesthesiol. 2002;19:447–451. 22. Chin KJ, Perlas A, Singh M, et al. An ultrasound-assisted approach 2. Harrison DA, Langham BT. Spinal anaesthesia for urologic surgery. A facilitates spinal anesthesia for total joint arthroplasty. Can J Anesth.2009; survey of failure rate, postdural puncture headache and patient satisfaction. 56:643–650. Anaesthesia. 1992;47:902–903. 23. Cork RC, Kryc JJ, Vaughan RW. Ultrasonic localization of the lumbar 3. Horlocker TT, McGregor DG, Matsushige DK, Schroeder DR, Besse JA. epidural space. Anesthesiology. 1980;52:513–516. A retrospective review of 4767 consecutive spinal anesthetics: central 24. Currie JM. Measurement of the depth to the extradural space using nervous system complications. Perioperative Outcomes Group. ultrasound. Br J Anaesth. 1984;56:345–347. Anesth Analg. 1997;84:578–584. 25. Ferre RM, Sweeney TW. Emergency physicians can easily obtain 4. Wulf H. Epidural anaesthesia and spinal hematoma. Can J Anaesth.1996; ultrasound images of anatomical landmarks relevant to lumbar puncture. 43:1260–1271. Am J Emerg Med. 2007;25:291–296. 5. Whitting PF,Rutjes AWS, Westwood ME, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 26. Gnaho A, Nguyen V,Villevielle T, Frota M, Marret E, Gentili ME. 2013;155:529–536. Assessing the depth of the subarachnoid space by ultrasound. Rev Bras Anestesiol. 2012;2:520–530. 6. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials.1996; 27. Grau T, Leipold RW,Conradi R, Martin E. Ultrasound control for presumed – 17:1–12. difficult epidural puncture. Acta Anaesthesiol Scand. 2001;45:766 771. 7. Higgins JPT, Altman DG, Gotzsche PC, et al. The Cochrane 28. Grau T, Leipold RW, Conradi R, Martin E, Motsch J. Ultrasound imaging Collaboration's tool for assessing risk of bias in randomized trials. BMJ. facilitates localization of the epidural space during combined spinal and – 2011;343:d5928. epidural anesthesia. Reg Anesth Pain Med.2001;26:64 67. 8. US Department of Health and Human Services Agency for Health Care 29. Grau T, Leipold RW, Conradi R, Martin E, Motsch J. Efficacy of Policy and Research. Acute Pain Management: Operative or Medical ultrasound imaging in obstetric epidural anesthesia. JClinAnesth. 2002; Procedures and Trauma. Clinical Practice Guideline I.Rockville,MD: 14:169–175. AHCPR; 1993. AHCPR Publication 92-0023-0107. 30. Helayel PE, da Conceicao DB, Meurer G, Swarovsky C, de Oliveira 9. Moher D, Liberati A, Tetzalff J, Altman DG, and the PRISMA Group. Filho GR. Evaluating the depth of the epidural space with the use of Preferred reporting items for systematic reviews and meta-analyses: the ultrasound. Rev Bras Anestesiol.2010;60:376–382. – PRISMA statement. Ann Intern Med. 2009;151:264 269. 31. Tran D, Kamani AA, Lessoway VA, Peterson C, Hor KW, Rohling RN. 10. Shaikh F, Brzezinski J, Alexander S, et al. Ultrasound imaging for lumbar Preinsertion paramedian ultrasound guidance for epidural anesthesia. punctures and epidural catheterisations: systematic review and Anesth Analg. 2009;109:661–667. meta-analysis. BMJ. 2013;346:f1720. 32. Vallejo MC, Phelps AL, Singh S, Orebaugh SL, Sah N. Ultrasound 11. Furness G, Reilly MP, Kuchi S. An evaluation of ultrasound imaging for decreases the failed labor epidural rate in resident trainees. Int J Obstet identification of lumbar intervertebral level. Anaesthesia. 2002;57:277–280. Anesth. 2010;19:373–378.

33. Chin KJ, Perlas A, Chan V,Brown-Shreves D, Koshkin A, Vaishnav V. 40. Mofidi M, Mohammadi M, Saidi H, et al. Ultrasound guided lumbar Ultrasound imaging facilitates spinal anesthesia in adults with difficult puncture in emergency department: time saving and less complications. surface anatomic landmarks. Anesthesiology. 2011;115:94–101. JResMedSci. 2013;18:303–307. 34. Grau T, Leipold RW, Fatehi S, Martin E, Motsch J. Real-time ultrasonic 41. Peterson MA, Pisupati D, Heyming TW, Abele JA, Lewis RJ. observation of combined spinal-epidural anaesthesia. Eur J Anaesthesiol. Ultrasound for routine lumbar puncture. Acad Emerg Med. 2014;21: 2004;21:25–31. 130–136. 35. Nomura JT, Leech SJ, Shenbagamurthi S, et al. A randomized controlled 42. Sahin T, Balaban O, Sahin L, Solak M, Toker K. A randomized trial of ultrasound-assisted lumbar puncture. JUltrasoundMed. 2007;26: controlled trial of preinsertion ultrasound guidance for spinal 1341–1348. anesthesia in pregnancy: outcomes among obese and lean parturients: 36. Wang Q, Yin C, Wang TL. Ultrasound facilitates identification of ultrasound for spinal anesthesia in pregnancy. JAnesth. 2014;28: – combined spinal-epidural puncture in obese parturients. Chin Med J (Engl). 413 419. 2012;125:3840–3843. 43. Perlas A. Evidence for the use of ultrasound in neuraxial blocks. – 37. Abdelhamid SA, Mansour AM. Ultrasound-guided intrathecal anesthesia: RegAnesthPainMed. 2010;35:S43 S46. does scanning help? Egypt J Anesth. 2013;29:389–394. 44. Moen V,Dahlgren N, Irestedt L. Severe neurological complications 38. Ansari T, Yousef A, El Gamassy A, Fayez M. Ultrasound-guided spinal after central neuraxial blockades in Sweden 1990-1999. Anesthesiology. – anesthesia in obstetrics: is there an advantage over the landmark 2004;101:950 959. technique in patients with easily palpable spines? Int J Obstet Anesth. 45. Hamandi K, Mottershead J, Lewis T, Ormerod IC, Ferguson IT. 2014;23:213–216. Irreversible damage to the spinal cord following spinal anesthesia. – 39. Lim YC, Choo CY, Tan KT. A randomized controlled trial of Neurology. 2002;59:624 626. ultrasound-assisted spinal anesthesia. Anaesth Intensive Care. 2014;42: 46. Reynolds F. Damage to the conus medullaris following spinal anaesthesia. 191–198. Anaesthesia. 2001;56:238–247.